Hoping someone can help figure out this mystery.

I installed build 90 on a fresh server last Monday 08/23. All was fine with good feedback from users, I have 13 users testing. On Sunday morning about 2am 08/29 the GW Administrator upgraded all sites to 8.02.

Since 2:50am 08/23 default.pipeline1.groupwise-AppInterface.log shows connection attempts for user stideswe. These fail causing lots of the following SOAP threads
POA soap thread details:
GW UID stideswe
User Agent Python-urllib/2.6
User IP (Synchroniser server IP)
Access Mode SOAP Sessions
Current Transaction getItemRequest
Time Elapsed (Secs) 0

Physical sessions owned and numbers of sessions and threads in total stay relatively low <10 this time around.

This caused the cpu load to quadruple. On Monday morning when users start coming in and started working, the load got higher and the administrator was alerted. When I stopped the datasync service at 10:30am 08/30 the load went back down instantly. When I started the services again around 11:30am the load stayed normal. And Stideswe no longer appeared in the logs.

So what's the mystery? Well stideswe is not a Novell Data Synchronizer user. He is the groupwise administrator. He was on the system once about a month ago, added and removed via consoleone group within two weeks. Removal was prompted when the data synchronizer was upgraded and his sync failed and caused a similar soap load issue. The device he used was set up at least three other times for other users before I wiped it on Mon 11:45am 08/30. He is not in the datasync database, he is not in the groupwise or mobility connector, he is not in the consoleone groups. Users have been added to the groups since he was removed. My process for upgrading is as follows: I have two servers. So I go to the one not in use, wipe it (this time I had new HDs,) load sles 11 from disc, upgrade, install mobility pack. Add the groups after installation not during. Let the devices authenticate in.

There are two sides to this problem one being GW 8 SOAP handling for stideswe the other being why did groupwise connector start trying to synchronize stideswe. And the latter is what I am most interested in at the moment, Novell support have given us some code to install for groupwise but either way I can accept bugs with SOAP that are being taken care of. Unpredictable behaviour is another thing.

Thought I was in the clear and would never get to test the code anyway after the services restarted on Mon 08/30 11:30am and stideswe was nowhere to be found in the groupwise-AppInterface log. However this afternoon 08/31 5:28pm stideswe appeared again. The load increased, this time Stideswe owned 120 physical SOAP sessions, PO crashed:

Aug 31 17:38:37 gwten01 kernel: gwpoa[14159]: segfault at 000000009e80888f rip 00000000f740e39a rsp 00000000f09eaff0 error 4

As soon as I turned off data synchronizer services load returned to normal.

Please help me get rid of stideswe. I know that the datasync database will only show users if they are in the mobility connector. Is there any way to see the groupwise info? Does anyone have any idea why these connection requests suddenly appear?

Logs to follow.

Any insight would be appreciated.